Is copy trading profitable? What the numbers say

The honest answer

Copy trading is profitable when three things are true at once: the lead's strategy has a real edge net of costs, your copy reproduces it faithfully, and you stay in through the drawdowns. Most copy-trading losses come from the second and third, not the first. A copier can lose money copying a strategy that made money.

Subtraction 1: the lead's edge, net of costs

Start with what the strategy makes after its own trading costs and funding, over a period long enough to mean something. For our book, that is a backtest of 41.8% a year over 2021–26, with a −18.8% worst drawdown. The costs inside that figure are about 3% of equity a year in fees and slippage and 2.7% in funding: a 5.7-point hurdle before the strategy has earned anything. A lead trader whose gross return is 8% a year on a similar turnover is, net, roughly flat.

Subtraction 2: the profit share

Binance's default profit share on public portfolios is 10% of your profit; Hyperliquid user vaults charge 10% per their published terms. It is charged on profit only, so it scales with the good years: a 30% year costs 3 points, a losing year costs nothing. Over a full cycle at a 20–40% target it is worth budgeting 2–4 points a year.

Subtraction 3: drift

Your copy is only as good as its ability to place the lead's orders. On Binance every order must clear a minimum value, and a systematic book sends many small ones. Sized against Binance's minimums, our backtest's orders were unplaceable 68% of the time for a $100 copy, 7% for $1,000 and 1% for $2,500. Each skipped order leaves your position a different size from the lead's, and the gap compounds. Below about $1,000, the copy is a different strategy from the one you chose.

Subtraction 4: behaviour

This is the largest and the least discussed. Over 300 weeks of backtest, the book was negative in 52% of weeks, 42% of four-week periods and 25% of three-month periods, while compounding at 41.8% a year. A copier who checks weekly sees red more often than green. The common response is to stop copying during a drawdown and restart after a recovery, which converts a strategy that made money into a personal record that did not.

Backtest, net of costs and funding, every overlapping window. Hypothetical.

A worked year

Suppose a copier allocates $2,500 to a strategy that returns 30% net of its own costs in a given year. Profit share at 10% takes 3 points: 27%. Drift at $2,500 is negligible. If the copier holds the full year, they make about $675. If they stop copying at the −6% three-month mark and restart two months later, they typically miss the recovery burst that trend strategies produce, and end the year near flat. Same strategy, same year, two outcomes; the difference was the copier.

So, is it profitable?

For a copier who chooses a lead by drawdown and record length rather than ROI, copies with enough capital to avoid drift, sets a fixed-ratio copy with a wide stop, and then leaves it alone for a year: yes, to roughly the lead's net return minus the profit share. For everyone else, the platform, the exchange and the lead are paid and the copier is not. The strategy's job is the first subtraction. The other three are yours.